Probability-Based Synthetic Minority Oversampling Technique

Many real-life datasets suffer from class imbalance, where one or more classes are under-represented in the dataset, resulting reduced classifier performance, with expected decline quality of procedures depending on classification results, such as financial losses to businesses inferior product quality. Improving accuracy by handling imbalance will positively impact accuracy. In this study, we present a Probability-Based Synthetic Minority Oversampling Technique (P-SMOTE) generate new examples for minority class. Our proposed solution improves enhancing oversampled through sampling probability distributions data. Results show improved performance over algorithms literature, an average F-score 0.821 13 using 5 classifiers.